The project aims to explore a novel approach, called system-subsystem dependency network, for modeling large, complex systems. This pilot study is primarily motivated by an application of system analysis to the study of disablement in older adults. Like many other system approaches to social and behavioral sciences, the modeling of disablement involves diverse data sources such as pilot experiments, epidemiologic and clinical studies, census data, and surveillance systems. It also involves multiple subsystems - components within a system that are individually cohesive both in structures and functions. While subsystem provides a meaningful overlay of external and tacit knowledge over the entire system, they often have to be individually calibrated using diverse data sources and different data sets. A methodological challenge is to reconcile overlapping subsystems as well as the potential parametric inconsistency across individually calibrated subsystems. The project extends a flexible tool - the generalized dependency network - for solving such problems. Through simulation experiments, state-of-the-art computational tools are tested for implementing the system- subsystem dependency network. Besides exploring the key idea of using generalized dependency network for operationalizing the system-subsystem approach, this pilot project also empirically tests the applicability of the method to a simplified disablemet model, which is restricted to the physical and cognitive subsystems, using data from two studies of aging. PUBLIC HEALTH RELEVANCE: Several methodological challenges hinder advance of the system science approach. One of them is the existence of fragmented data - i.e., a comprehensive data set covering all aspects of the system is usually not available. The project uses an innovative approach, called the system-subsystem dependency network, which is capable of integrating subsystems that have been individually calibrated using separate data sets.